Kwang-Hyun Kim Defied Odds in 2020. How and Can It Continue?

Fellow Birds on the Black contributor, Tara Wellman, was recently on another podcast and, at the very end of what I thought was a fantastic job running down the Cardinals line up construction and story lines coming into the 2021 season, she talked a bit about 2020 veteran/rookie phenom Kwang-Hyun Kim. She gave her thoughts on why she believes KK, as #STLCards fans affectionately call him, might have had the success he did in 2020. This thought fascinated me and while Tara was talking over the speakers of my laptop, I began to dive down the Baseball Savant rabbit hole of information on KK.

Before typing anything up today, I did a cursory glance (at best) over Cardinals blogs to see who had written what about KK so that I wasn't re-inventing the wheel here. Over at Viva El Birdos, Blake Newberry already took a look at a possible reason as to why KK was so effective in 2020's short sample, despite not having the best "stuff" in the league as a pitcher. That was a few months after Blake wondered if Kim could be effective again. I think that Blake's work was a good starting point, but I wanted to dive deeper. C70 also listed a couple of things that Kim did statistically well and poorly in 2020. Those things were where I began.

The first thing I wanted to take a look at today was the pitch distribution for Kwang-Hyun Kim in 2020. If this distribution holds for the future, KK will throw:

  • 49% four-seam fastballs that average 90 mph

  • 31% sliders that average 83 mph

  • 11% curveballs that average 70 mph

  • 8% change ups that average 80 mph

The other thing to notice in that graphic above, from Baseball Savant, is the dotted lines that represent league average. His four-seam lives a tick below league average, as you can see in the distribution of pitches of his. His slider is basically league average speed. However, his change up barely ever reaches league average - like the fastball - which gives it the proper mph separation to be a literal "change up" from his fastball. His curve is one that would be considered a "slow curve" for some - with it averaging around 70 mph but being a right-tail distribution, most of them seem to be 70 mph or even slower. The 76 or 77 mph outliers drag that average up higher than where the curveball seems to sit. Batters hit just .247/.286/.416/.702 on pitches 70 miles per hour or slower in 2020, league-wide. That was the best players have hit on that slow of a pitch since 2015. Hitters today are geared up for 95+ mph. A change in velocity that great can throw hitters off a bit.

The above chart shows where Kwang-Hyun Kim earned or lost runs based on where he threw the ball compared to the strike zone. Baseball Savant breaks these zones down into four regions. They explain it better than I can:

On this chart, negative numbers are GOOD and positive numbers are BAD because pitchers want to PREVENT runs, not allow them. So KK did quite well, having a -9 run value. That means that Kwang-Hyun Kim was 9 runs better than league average on his pitches around the zone and we can see where he excelled.

Kim only lost 2 runs from waste pitches and 2 runs from chase pitches. Where Kim did incredibly well was throwing 3% less than league average pitches in the heart of the zone and getting 2% less swings on the ones he did throw there. That's really good because those pitches are the easiest to hit; this shows that he either knew when to throw strikes or he was deceptive. With what is to come in this analysis, I believe he's got some hidden deception somewhere within his delivery. He also threw 4% more pitches in the shadow zone than league average, living in that region over the plate that is quite questionable for the hitter. Now, Kim did get less swings and more takes in that shadow zone than league average, however. He threw more pitches there, so maybe those overall numbers of swings averaged out, however.

Now, Baseball Savant doesn't have a pretty chart for where he got or lost strikes on individual pitches (fastball, slider, curveball, change up), but we can see WHERE he threw individual pitches and I can do the legwork to break down how many of each were where for you. First, here is a look at heatmaps of each pitch type from Kim in 2020, followed by where individual pitches landed:

You can see that his four-seam fastball seemed to live on the shadow of the zone while his slider and curveball found the heart too often for my liking (unless they either got takes or missed bats) and the change up lived mostly out of the zone in the chase region. Narrator: But the curveball did not miss enough bats to be thrown where it was.

The first (left) table shows how many pitches of each type he threw in each region.

The second (middle) table shows what percent of all of his pitches went into each region.

The third (right) table shows what percent of each individual pitch went into each region.

You can see from those tables - especially the last one - where he lived most often with each individual pitch. Notice that KK wasted a lot of curveballs, but also put a lot of them in the shadow of the zone. Here's one example of his too many wasted curveballs. Now, one good thing about this is, he never threw a pitch in the waste zone when he had 2 or 3 balls already on a hitter. They were all thrown when the count started with a 0 or a 1. That's better, I suppose.

In fact, Kim's curveball either seemed to find it's way there or seemed to find the bat in a bad way. One example of that is about where the ball was hit on average when a player made contact. This one was a 71 mph variety right down the heart of the zone. You can't "miss" there. Nope, nope, nope.

Also notice above how Kim attempted to get chases with his change up quite often, but probably didn't live close enough to the shadow zone on it often enough. But this article isn't about how Kim was unsuccessful. Because in 2020, Kim's results were quite successful.

Kim was very effective at nibbling the corners with his two main pitches - his fastball and his slider. He rarely wasted those two pitches at all. Let's look closer at those. In fact, here are the called strikes on fastballs and sliders that were in the shadow of the zone - an overlay of the heat map of those two types of pitches and the actual locations themselves.

I would also like to show you what some of the contact looked like on those shadow fastballs and shadow sliders - nearly 50% of which were batted at under 80 miles per hour.

Kwang-Hyun Kim only gave up 5 barreled baseballs on 613 pitches in 2020.

Honestly, if people can barrel those two balls on the left-handed batter's box side of the plate, good on them. That slider was away to a lefty (one of the pitches in the red dot). But all in all, THAT'S IT on the barrels - for the year! He did not give up super hard contact either. In fact, below are the only 5 pitches in which he gave up even "solid contact" that was 95 mph or harder off of the bat.

Even those pitches had two to three that were not bad pitches at all. Two of them were probably ones that he would want back, though.

Those are the stats that C70 entered into his article I linked to earlier. KK's barrel rate of 4.2% was just 65% of league average. KK's solid contact rate of 4.2% was just 75% of league average. The number of flares or burners that KK allowed were right about league average, as were the number of topped grounders he allowed. So where did those extra batted balls end up? KK's 2020 short season had 217% the league average of "weak"ly hit balls. He also had 5% more than league average balls that were hit under the ball - pitches that were popped up. All in all, OVER 2/3 of all batted balls off of KK were weakly hit, popped up, or topped into the ground. That's really, really good. His percentage of those landed at 66.9% - when league average for those three types of batted balls is 61.3%. That's 9% better than league average.

Those numbers are how KK can have a strikeout percentage that's in the 8th percentile and still be quite effective. He was 71st percentile in the exit velocity of all batted balls against him. Why? He was 84th percentile in both hard hit balls and barreled balls against - so while his FIP and xFIP do not lend themselves towards being favorable for KK's future, his xERA - Baseball Savant's version of FIP - says that he was above league average last year. Basically, while he wasn't going to maintain a 1.6 ish ERA doing what he was doing, he should have an above league average ERA if he pitches like he did last opposed to what his 4.52 xFIP and 4.17 (steamer) or 4.34 (zips) projections might suggest.

So can KK's success continue?

Well, above we talked about if he can pitch like last year, then yes. He should still be better than league average at allowing runs. Let's look at another piece here, though.

Baseball Savant has a tool called a pitcher (or hitter) similarity tool. You can actually compare pitchers two different ways. If you compare pitchers by hitting profile (what I outlined above with barreled balls, solid hit balls, weakly hit balls, etc.) two of the most similar pitchers to Kim's 2020 season were Adam Wainwright's 2020 season and Kyle Hendricks' 2020 season. That's fantastic for Kim, as those pitchers also did quite well. If you compare pitchers by speed and movement on their pitches, Kim compares to Nick Margevicius the most, but then Mike Minor, Clayton Kershaw, and Daniel Castano (who we traded away in the Marcell Ozuna deal). Let's look at those 6 players.

  • Kershaw had a .251 wOBA allowed and 3.06 xERA last year

  • Hendricks had a .269 wOBA allowed and 3.13 xERA last year

  • Wainwright had a .277 wOBA allowed and 4.52 xERA last year

  • Minor had a .307 wOBA allowed and 4.33 xERA last year

  • Margevicius had a .314 wOBA allowed and 4.36 xERA last year

  • Castano had a .317 wOBA allowed and 5.24 xERA last year

I'll let you look through that list and see who you think he'll be most like in 2020, but his numbers last year were a .251 wOBA allowed and 3.85 xERA last year. I've got my suppositions, but I'll put those in my player projection later this month.